Feature request: continuous or interval-based recognition

[daltux]

It would be very nice to have an option to do continous recognition or, if more viable, to start it every minute (or some user setting).

[aleksey-saenko]

Hello, thanks for the suggestion.

This feature would be useful, and I have already thought about implementing it. It's possible to optimize the number of requests to the service: we can avoid sending during silence, we can make requests only after the end of the track (the app knows the track duration and the timestamp when the recognition occurred). However, the number of requests will still be large, and this is a problem.

AudD is not a free service and requires payment for requests. It is assumed that the developer buys a token and provides a paid application or monetizes it in another way, such as through advertising. I created this application not for commercial purposes, so the application relies on the user to obtain a token.

I think most users rely on current limit, which is around 3-5 recognitions per day (the new token has more requests, but only for 2 weeks). Therefore, as long as the app exists in this form, I don't think this feature will be in high demand.

I will leave this issue open, as perhaps a solution will be found someday.